McKinsey: agentic AI adoption soars, but ROI stalls

🕒 Published on Zendoric: September 4, 2026 · 09:12
✨ AI-generated · how it's made
The email presents the latest findings from McKinsey's "state of AI research" on enterprise adoption of artificial intelligence. According to the data cited, companies are integrating AI ever more deeply into their business functions, but McKinsey raises the central question of the newsletter: is that…
By McKinsey & Company (Dan Tinkoff and Lieven Van der Veken, global leaders of QuantumBlack, AI by McKinsey).
The email presents the latest findings from McKinsey's "state of AI research" on enterprise adoption of artificial intelligence. According to the data cited, companies are embedding AI ever more deeply into their business functions, but McKinsey raises the email's central question: is that investment actually translating into financial results?
The specific figures the email provides are as follows: 40 percent of large companies are already scaling autonomous agents, up from 27 percent the previous year. In addition, close to a third of these organizations are building software in-house using agentic coding tools. Yet despite this growth in adoption, enterprise-level EBIT impact has remained flat at 37 percent.
The email, signed by Dan Tinkoff and Lieven Van der Veken along with Michael Chui and Tara Balakrishnan as the research authors, poses as its key question what the organizations that do capture significant value from AI are doing differently, inviting readers to explore the full report under the title "Turn AI adoption into enterprise value".
Alongside the main article, the email links to several related pieces within the McKinsey newsletter: "Agents, robots, and us: Skill partnerships in the age of AI", which addresses how AI is expanding the productivity frontier and the need to rethink how people work alongside intelligent machines; "AI-powered software development: How technology is rewriting the rules", in which three McKinsey experts explain how AI is changing software creation, from individual productivity tricks to the complete reinvention of workflows and roles; "The cost of intelligence: How CIOs can manage AI demand at scale", on how CIOs must manage AI demand at enterprise scale amid rising costs, optimizing for outcomes and not just cost; and "From adoption to impact: Three horizons of AI transformation", based on a global survey describing how organizations can progress from individual adoption to value capture across the entire enterprise.
Taken together, the email's editorial message is that, although adoption of autonomous agents and agentic coding tools is growing strongly, most companies still fail to turn that use into measurable, sustained impact on the bottom line, and McKinsey positions its research as a guide to closing that gap between adoption and value.
🔗 Related on Zendoric


